SegCLIP: Patch Aggregation with Learnable Centers for Open-Vocabulary Semantic Segmentation
About
Recently, the contrastive language-image pre-training, e.g., CLIP, has demonstrated promising results on various downstream tasks. The pre-trained model can capture enriched visual concepts for images by learning from a large scale of text-image data. However, transferring the learned visual knowledge to open-vocabulary semantic segmentation is still under-explored. In this paper, we propose a CLIP-based model named SegCLIP for the topic of open-vocabulary segmentation in an annotation-free manner. The SegCLIP achieves segmentation based on ViT and the main idea is to gather patches with learnable centers to semantic regions through training on text-image pairs. The gathering operation can dynamically capture the semantic groups, which can be used to generate the final segmentation results. We further propose a reconstruction loss on masked patches and a superpixel-based KL loss with pseudo-labels to enhance the visual representation. Experimental results show that our model achieves comparable or superior segmentation accuracy on the PASCAL VOC 2012 (+0.3% mIoU), PASCAL Context (+2.3% mIoU), and COCO (+2.2% mIoU) compared with baselines. We release the code at https://github.com/ArrowLuo/SegCLIP.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Semantic segmentation | PASCAL VOC 2012 (val) | Mean IoU52.6 | 2142 | |
| Semantic segmentation | ADE20K | mIoU8.7 | 1024 | |
| Semantic segmentation | PASCAL VOC (val) | mIoU52.6 | 362 | |
| Semantic segmentation | PASCAL Context (val) | mIoU24.7 | 360 | |
| Semantic segmentation | Pascal VOC (test) | mIoU52.6 | 236 | |
| Semantic segmentation | Pascal Context | mIoU24.7 | 217 | |
| Semantic segmentation | Pascal Context (test) | mIoU24.7 | 191 | |
| Semantic segmentation | Pascal VOC | mIoU0.526 | 180 | |
| Semantic segmentation | COCO (val) | mIoU26.5 | 150 | |
| Semantic segmentation | Pascal Context 60 | mIoU24.7 | 139 |